Fixing lack of ref_model

data/code_eval_board.csv

@@ -8,6 +8,7 @@ T,Models,ARC,HellaSwag,MMLU,TruthfulQA,Winogrande,GSM8K,Reference Model
 🔶,upstage/SOLAR-10.7B-Instruct-v1.0,0.11,0.49,0.28,0.36,0.01,0.96,huggyllama/llama-7b
 🔶,AIDC-ai-business/Marcoroni-7B-v3,0.1,0.14,0.2,0.41,0.0,0.95,mistralai/Mistral-7B-v0.1
 🔶,amazon/MistralLite,0.09,0.14,0.2,0.43,0.0,0.73,mistralai/Mistral-7B-v0.1
-🔶,openchat/openchat_3.5,0.13,0.13,0.23,0.45,0.0,0.97
-🔶,meta-math/MetaMath-Mistral-7B,0.08,0.1,0.17,0.42,0.0,0.97
-🔶,teknium/OpenHermes-2.5-Mistral-7B,0.07,0.13,0.23,0.39,0.0,0.96
+🔶,openchat/openchat_3.5,0.13,0.13,0.23,0.45,0.0,0.97,mistralai/Mistral-7B-v0.1
+🔶,meta-math/MetaMath-Mistral-7B,0.08,0.1,0.17,0.42,0.0,0.97,mistralai/Mistral-7B-v0.1
+🔶,teknium/OpenHermes-2.5-Mistral-7B,0.07,0.13,0.23,0.39,0.0,0.96,mistralai/Mistral-7B-v0.1
+